Validating a Markov model of treatment for hepatitis C virus-related hepatocellular carcinoma.

Objective: We created and validated a Markov model to simulate the prognosis with treatment for HCV-related hepatocellular carcinoma (HCC) for assessment of cost-effectiveness for alternative treatments of HCC.
Method: Markov state incorporated into the model consisted of the treatment as a surrogate for HCC stage and underlying liver function. Retrospective data of 793 patients from three university hospitals were used to determine Kaplan-Meier survival curves for each treatment and transition probabilities were derived from them.
Results: There was substantial overlap in the 95% CIs of the Markov model predicted and the Kaplan-Meier survival curves for each therapy. The predicted survival curves were also similar with those from the nationwide survey data supporting the external validity of our model.
Conclusions: Our Markov model estimates for prognosis with HCC have both internal and external validity and should be considered applicable for estimating cost-effectiveness related to HCC.